Article
Welcome Inkling by Thinking Machines
- +3
burtenshaw, merve, pcuenq, ariG23498, andito
• • 141
Cem Bıkmaz
ceofast
AI & ML interests
Computer Vision, Natural Language Processing, Large Language Models, Data Science, MLOps, Data Engineering
Recent Activity
upvoted an article about 5 hours ago
Welcome Inkling by Thinking Machines updated a model 10 months ago
ceofast/kumru-2b-lora published a model 10 months ago
ceofast/kumru-2b-loraOrganizations
upvoted an article about 5 hours ago
updated a
model 10 months ago
published a
model 10 months ago
upvoted 5
changelogs 10 months ago
Hugging Face Changelog
Introducing HF Jobs: Run scalable compute jobs on Hugging Face
• 207
Hugging Face Changelog
Emoji Autocomplete in Discussions and Posts
• 72
Hugging Face Changelog
Switch settings context between user and organizations
• 45
Hugging Face Changelog
Custom Domains for Spaces
• 89
Hugging Face Changelog
Repositories total file size is now displayed
• 179
liked a
Space 10 months ago
updated a
model about 1 year ago
published a
model about 1 year ago
updated a
model almost 2 years ago
liked a
model almost 2 years ago
updated a
model about 2 years ago
liked a
model about 2 years ago
reacted to merve's post with 🚀 about 2 years ago
Post
2063
Do we fully leverage ViT encoders in vision language models?
A new paper (by @HuanjinYao et al) built a dense connector that does it better! HuanjinYao/DenseConnector-v1.5-8B
HuanjinYao/denseconnector-66500e173fc8c9f05dc98dea
VLMs consist of an image encoder block, a projection layer that projects image embeddings to text embedding space and then a text decoder sequentially connected 📖
This paper explores using intermediate states of image encoder and not a single output 🤩
The authors explore three different ways of instantiating dense connector: sparse token integration, sparse channel integration and dense channel integration. (see paper on how they do it Dense Connector for MLLMs (2405.13800))
They explore all three of them integrated to LLaVA 1.5 and found out each of the new models are superior to the original LLaVA 1.5 🥹 I tried the model and it seems to work very well. As part of the release, the authors have released various ckpts based on different decoders (Vicuna 7/13B and Llama 3-8B) that you can find in the collection 🤗
A new paper (by @HuanjinYao et al) built a dense connector that does it better! HuanjinYao/DenseConnector-v1.5-8B
HuanjinYao/denseconnector-66500e173fc8c9f05dc98dea
VLMs consist of an image encoder block, a projection layer that projects image embeddings to text embedding space and then a text decoder sequentially connected 📖
This paper explores using intermediate states of image encoder and not a single output 🤩
The authors explore three different ways of instantiating dense connector: sparse token integration, sparse channel integration and dense channel integration. (see paper on how they do it Dense Connector for MLLMs (2405.13800))
They explore all three of them integrated to LLaVA 1.5 and found out each of the new models are superior to the original LLaVA 1.5 🥹 I tried the model and it seems to work very well. As part of the release, the authors have released various ckpts based on different decoders (Vicuna 7/13B and Llama 3-8B) that you can find in the collection 🤗
upvoted a paper about 2 years ago
liked a
model about 2 years ago
updated 2
models over 2 years ago